Add 64-bit slm optimized strlcpy and srlcat.

Change-Id: Ic948934d91c83bbfdfd00c05ee8b14952e012549
Signed-off-by: Varvara Rainchik <varvara.rainchik@intel.com>
diff --git a/libc/arch-x86_64/string/sse2-strlcpy-slm.S b/libc/arch-x86_64/string/sse2-strlcpy-slm.S
new file mode 100755
index 0000000..9d4b52f
--- /dev/null
+++ b/libc/arch-x86_64/string/sse2-strlcpy-slm.S
@@ -0,0 +1,1062 @@
+/*
+Copyright (c) 2014, Intel Corporation
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without
+modification, are permitted provided that the following conditions are met:
+
+    * Redistributions of source code must retain the above copyright notice,
+    * this list of conditions and the following disclaimer.
+
+    * Redistributions in binary form must reproduce the above copyright notice,
+    * this list of conditions and the following disclaimer in the documentation
+    * and/or other materials provided with the distribution.
+
+    * Neither the name of Intel Corporation nor the names of its contributors
+    * may be used to endorse or promote products derived from this software
+    * without specific prior written permission.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
+ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+(INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
+ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+*/
+
+#ifndef L
+# define L(label)	.L##label
+#endif
+
+#ifndef cfi_startproc
+# define cfi_startproc	.cfi_startproc
+#endif
+
+#ifndef cfi_endproc
+# define cfi_endproc	.cfi_endproc
+#endif
+
+#ifndef ENTRY
+# define ENTRY(name)	\
+	.type name, @function;	\
+	.globl name;	\
+	.p2align 4;	\
+name:	\
+	cfi_startproc
+#endif
+
+#ifndef END
+# define END(name)	\
+       cfi_endproc;	\
+       .size name, .-name
+#endif
+
+
+#ifndef STRLCPY
+# define STRLCPY	strlcpy
+#endif
+
+#define JMPTBL(I, B)	I - B
+#define BRANCH_TO_JMPTBL_ENTRY(TABLE, INDEX, SCALE)	\
+	lea	TABLE(%rip), %r11;	\
+	movslq	(%r11, INDEX, SCALE), %rcx;	\
+	lea	(%r11, %rcx), %rcx;	\
+	jmp	*%rcx
+
+#define RETURN	\
+	add	%r9, %rax;	\
+	ret
+
+.text
+ENTRY (STRLCPY)
+	xor	%rax, %rax
+	xor	%r9, %r9
+	mov	%rdx, %r8
+	cmp	$0, %r8
+	jz	L(CalculateSrcLen)
+
+#ifdef USE_AS_STRLCAT
+	xor	%rcx, %rcx
+	pxor	%xmm0, %xmm0
+
+	movdqu	(%rdi), %xmm1
+	pcmpeqb %xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+
+	cmp	$17, %r8
+	jb	L(SizeEndCase1)
+	test	%rdx, %rdx
+	jnz	L(StringEndCase1)
+
+	add	$16, %rax
+	movdqu	16(%rdi), %xmm1
+	pcmpeqb %xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+
+	cmp	$33, %r8
+	jb	L(SizeEndCase1)
+	test	%rdx, %rdx
+	jnz	L(StringEndCase1)
+
+	mov	%rdi, %rcx
+	and	$15, %rcx
+	and	$-16, %rdi
+
+	add	%rcx, %r8
+	sub	$16, %r8
+
+L(DstLenLoop):
+	movdqa	(%rdi, %rax), %xmm1
+	pcmpeqb %xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+	sub	$16, %r8
+	jbe	L(SizeEndCase2)
+	test	%rdx, %rdx
+	jnz	L(StringEndCase2)
+	add	$16, %rax
+	jmp	L(DstLenLoop)
+
+L(StringEndCase2):
+	add	$16, %r8
+	bsf	%rdx, %rdx
+	sub	%rdx, %r8
+	add	%rdx, %rax
+	sub	%rcx, %r9
+	add	%rax, %rdi
+	jmp	 L(CopySrcString)
+
+L(SizeEndCase1):
+	test	%rdx, %rdx
+	jz	L(SizeEnd)
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	cmp	%r8, %rax
+	jb	L(StringEnd)
+L(SizeEnd):
+	mov	%r8, %r9
+	jmp	L(CalculateSrcLenCase1)
+
+L(SizeEndCase2):
+	add	$16, %r8
+	test	%rdx, %rdx
+	jz	L(StringEndCase4)
+	bsf	%rdx, %rdx
+	cmp	%r8, %rdx
+	jb	L(StringEndCase3)
+L(StringEndCase4):
+	add	%r8, %rax
+	sub	%rcx, %rax
+	mov	%rax, %r9
+	jmp	L(CalculateSrcLenCase1)
+
+L(StringEndCase3):
+	add	%rdx, %rax
+	sub	%rcx, %r9
+	add	%rax, %rdi
+	sub	%rdx, %r8
+	jmp	L(CopySrcString)
+
+L(StringEndCase1):
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	sub	%rcx, %rax
+L(StringEnd):
+	add	%rax, %rdi
+	sub	%rax, %r8
+#endif
+
+	mov	%rsi, %rcx
+	and	$63, %rcx
+	cmp	$32, %rcx
+	jbe	L(CopySrcString)
+
+	and	$-16, %rsi
+	and	$15, %rcx
+	pxor	%xmm0, %xmm0
+	pxor	%xmm1, %xmm1
+
+	pcmpeqb	(%rsi), %xmm1
+	pmovmskb %xmm1, %rdx
+	shr	%cl, %rdx
+	mov	$16, %r10
+	sub	%rcx, %r10
+	cmp	%r10, %r8
+	jbe	L(CopyFrom1To16BytesTailCase2OrCase3)
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesTail)
+
+	pcmpeqb	16(%rsi), %xmm0
+	pmovmskb %xmm0, %rdx
+	add	$16, %r10
+	cmp	%r10, %r8
+	jbe	L(CopyFrom1To32BytesCase2OrCase3)
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To32Bytes)
+
+	movdqu	(%rsi, %rcx), %xmm1
+	movdqu	%xmm1, (%rdi)
+#ifdef USE_AS_STRLCAT
+	add	%rax, %r9
+#endif
+	jmp	L(LoopStart)
+
+	.p2align 4
+L(CopySrcString):
+#ifdef USE_AS_STRLCAT
+	add	%rax, %r9
+	xor	%rax, %rax
+#endif
+	pxor	%xmm0, %xmm0
+	movdqu	(%rsi), %xmm1
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+
+	cmp	$17, %r8
+	jb	L(CopyFrom1To16BytesTail1Case2OrCase3)
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesTail1)
+
+	movdqu	16(%rsi), %xmm2
+	pcmpeqb	%xmm2, %xmm0
+	movdqu	%xmm1, (%rdi)
+	pmovmskb %xmm0, %rdx
+	add	$16, %rax
+
+	cmp	$33, %r8
+	jb	L(CopyFrom1To32Bytes1Case2OrCase3)
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To32Bytes1)
+
+	mov	%rsi, %rcx
+	and	$15, %rcx
+	and	$-16, %rsi
+
+L(LoopStart):
+	sub	%rcx, %rdi
+	add	%rcx, %r8
+	sub	$16, %r8
+	mov	$16, %rax
+
+L(16Loop):
+	movdqa	(%rsi, %rax), %xmm1
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+	sub	$16, %r8
+	jbe	L(CopyFrom1To16BytesCase2OrCase3)
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesXmmExit)
+	movdqu	%xmm1, (%rdi, %rax)
+	add	$16, %rax
+	jmp	L(16Loop)
+
+/*------End of main part with loops---------------------*/
+
+/* Case1 */
+	.p2align 4
+L(CopyFrom1To16Bytes):
+	add	%rcx, %rdi
+	add	%rcx, %rsi
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+	.p2align 4
+L(CopyFrom1To16BytesTail):
+	add	%rcx, %rsi
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+	.p2align 4
+L(CopyFrom1To32Bytes1):
+	add	$16, %rsi
+	add	$16, %rdi
+	sub	$16, %r8
+L(CopyFrom1To16BytesTail1):
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+	.p2align 4
+L(CopyFrom1To32Bytes):
+	bsf	%rdx, %rdx
+	add	%rcx, %rsi
+	add	$16, %rdx
+	sub	%rcx, %rdx
+	add	%rdx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+	.p2align 4
+L(CopyFrom1To16BytesExit):
+	add	%rdx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+/* Case2 */
+
+	.p2align 4
+L(CopyFrom1To16BytesCase2):
+	add	$16, %r8
+	add	%rax, %rdi
+	add	%rax, %rsi
+	bsf	%rdx, %rdx
+	sub	%rcx, %rax
+	cmp	%r8, %rdx
+	jb	L(CopyFrom1To16BytesExit)
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To32BytesCase2):
+	add	%rcx, %rsi
+	bsf	%rdx, %rdx
+	add	$16, %rdx
+	sub	%rcx, %rdx
+	cmp	%r8, %rdx
+	jb	L(CopyFrom1To16BytesExit)
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+L(CopyFrom1To16BytesTailCase2):
+	add	%rcx, %rsi
+	bsf	%rdx, %rdx
+	cmp	%r8, %rdx
+	jb	L(CopyFrom1To16BytesExit)
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To16BytesTail1Case2):
+	bsf	%rdx, %rdx
+	cmp	%r8, %rdx
+	jb	L(CopyFrom1To16BytesExit)
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+/* Case2 or Case3,  Case3 */
+
+	.p2align 4
+L(CopyFrom1To16BytesCase2OrCase3):
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesCase2)
+	add	$16, %r8
+	add	%rax, %rdi
+	add	%rax, %rsi
+	add	%r8, %rax
+	sub	%rcx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To32BytesCase2OrCase3):
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To32BytesCase2)
+	add	%rcx, %rsi
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To16BytesTailCase2OrCase3):
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesTailCase2)
+	add	%rcx, %rsi
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To32Bytes1Case2OrCase3):
+	add	$16, %rdi
+	add	$16, %rsi
+	sub	$16, %r8
+L(CopyFrom1To16BytesTail1Case2OrCase3):
+	test	%rdx, %rdx
+	jnz	L(CopyFrom1To16BytesTail1Case2)
+	add	%r8, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitTable), %r8, 4)
+
+	.p2align 4
+L(CopyFrom1To16BytesXmmExit):
+	bsf	%rdx, %rdx
+	add	%rax, %rdi
+	add	%rax, %rsi
+	add	%rdx, %rax
+	sub	%rcx, %rax
+	BRANCH_TO_JMPTBL_ENTRY (L(ExitStringTailTable), %rdx, 4)
+
+/*------------End labels regarding with copying 1-16 bytes--and 1-32 bytes----*/
+
+
+	.p2align 4
+L(Exit0):
+	RETURN
+
+	.p2align 4
+L(Exit1):
+	movb	$0, (%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit2):
+	movb	(%rsi), %dh
+	movb	%dh, (%rdi)
+	movb	$0, 1(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit3):
+	movw	(%rsi), %dx
+	movw	%dx, (%rdi)
+	movb	$0, 2(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit4):
+	movw	(%rsi), %cx
+	movb	2(%rsi), %dh
+	movw	%cx, (%rdi)
+	movb	%dh, 2(%rdi)
+	movb	$0, 3(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit5):
+	movl	(%rsi), %edx
+	movl	%edx, (%rdi)
+	movb	$0, 4(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit6):
+	movl	(%rsi), %ecx
+	movb	4(%rsi), %dh
+	movl	%ecx, (%rdi)
+	movb	%dh, 4(%rdi)
+	movb	$0, 5(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit7):
+	movl	(%rsi), %ecx
+	movw	4(%rsi), %dx
+	movl	%ecx, (%rdi)
+	movw	%dx, 4(%rdi)
+	movb	$0, 6(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit8):
+	movl	(%rsi), %ecx
+	movl	3(%rsi), %edx
+	movl	%ecx, (%rdi)
+	movl	%edx, 3(%rdi)
+	movb	$0, 7(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit9):
+	movq	(%rsi), %rdx
+	movq	%rdx, (%rdi)
+	movb	$0, 8(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit10):
+	movq	(%rsi), %rcx
+	movb	8(%rsi), %dh
+	movq	%rcx, (%rdi)
+	movb	%dh, 8(%rdi)
+	movb	$0, 9(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit11):
+	movq	(%rsi), %rcx
+	movw	8(%rsi), %dx
+	movq	%rcx, (%rdi)
+	movw	%dx, 8(%rdi)
+	movb	$0, 10(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit12):
+	movq	(%rsi), %rcx
+	movl	7(%rsi), %edx
+	movq	%rcx, (%rdi)
+	movl	%edx, 7(%rdi)
+	movb	$0, 11(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit13):
+	movq	(%rsi), %rcx
+	movl	8(%rsi), %edx
+	movq	%rcx, (%rdi)
+	movl	%edx, 8(%rdi)
+	movb	$0, 12(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit14):
+	movq	(%rsi), %rcx
+	movq	5(%rsi), %rdx
+	movq	%rcx, (%rdi)
+	movq	%rdx, 5(%rdi)
+	movb	$0, 13(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit15):
+	movq	(%rsi), %rcx
+	movq	6(%rsi), %rdx
+	movq	%rcx, (%rdi)
+	movq	%rdx, 6(%rdi)
+	movb	$0, 14(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit16):
+	movq	(%rsi), %rcx
+	movq	7(%rsi), %rdx
+	movq	%rcx, (%rdi)
+	movq	%rdx, 7(%rdi)
+	movb	$0, 15(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit17):
+	movdqu	(%rsi), %xmm0
+	movdqu	%xmm0, (%rdi)
+	movb	$0, 16(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit18):
+	movdqu	(%rsi), %xmm0
+	movb	16(%rsi), %dh
+	movdqu	%xmm0, (%rdi)
+	movb	%dh, 16(%rdi)
+	movb	$0, 17(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit19):
+	movdqu	(%rsi), %xmm0
+	movw	16(%rsi), %cx
+	movdqu	%xmm0, (%rdi)
+	movw	%cx, 16(%rdi)
+	movb	$0, 18(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit20):
+	movdqu	(%rsi), %xmm0
+	movl	15(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	movl	%ecx, 15(%rdi)
+	movb	$0, 19(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit21):
+	movdqu	(%rsi), %xmm0
+	movl	16(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	movl	%ecx, 16(%rdi)
+	movb	$0, 20(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit22):
+	movdqu	(%rsi), %xmm0
+	movl	16(%rsi), %ecx
+	movb	20(%rsi), %dh
+	movdqu	%xmm0, (%rdi)
+	movl	%ecx, 16(%rdi)
+	movb	%dh, 20(%rdi)
+	movb	$0, 21(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit23):
+	movdqu	(%rsi), %xmm0
+	movq	14(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	movq	%rcx, 14(%rdi)
+	movb	$0, 22(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit24):
+	movdqu	(%rsi), %xmm0
+	movq	15(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	movq	%rcx, 15(%rdi)
+	movb	$0, 23(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit25):
+	movdqu	(%rsi), %xmm0
+	movq	16(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	movq	%rcx, 16(%rdi)
+	movb	$0, 24(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit26):
+	movdqu	(%rsi), %xmm0
+	movq	16(%rsi), %rcx
+	movb	24(%rsi), %dh
+	movdqu	%xmm0, (%rdi)
+	movq	%rcx, 16(%rdi)
+	mov	%dh, 24(%rdi)
+	movb	$0, 25(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit27):
+	movdqu	(%rsi), %xmm0
+	movq	16(%rsi), %rdx
+	movw	24(%rsi), %cx
+	movdqu	%xmm0, (%rdi)
+	movq	%rdx, 16(%rdi)
+	movw	%cx, 24(%rdi)
+	movb	$0, 26(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit28):
+	movdqu	(%rsi), %xmm0
+	movq	16(%rsi), %rdx
+	movl	23(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	movq	%rdx, 16(%rdi)
+	movl	%ecx, 23(%rdi)
+	movb	$0, 27(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit29):
+	movdqu	(%rsi), %xmm0
+	movq	16(%rsi), %rdx
+	movl	24(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	movq	%rdx, 16(%rdi)
+	movl	%ecx, 24(%rdi)
+	movb	$0, 28(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit30):
+	movdqu	(%rsi), %xmm0
+	movdqu	13(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 13(%rdi)
+	movb	$0, 29(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit31):
+	movdqu	(%rsi), %xmm0
+	movdqu	14(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 14(%rdi)
+	movb	$0, 30(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(Exit32):
+	movdqu	(%rsi), %xmm0
+	movdqu	15(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 15(%rdi)
+	movb	$0, 31(%rdi)
+	jmp	L(CalculateSrcLen)
+
+	.p2align 4
+L(StringTail0):
+	mov	(%rsi), %dl
+	mov	%dl, (%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail1):
+	mov	(%rsi), %dx
+	mov	%dx, (%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail2):
+	mov	(%rsi), %cx
+	mov	2(%rsi), %dl
+	mov	%cx, (%rdi)
+	mov	%dl, 2(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail3):
+	mov	(%rsi), %edx
+	mov	%edx, (%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail4):
+	mov	(%rsi), %ecx
+	mov	4(%rsi), %dl
+	mov	%ecx, (%rdi)
+	mov	%dl, 4(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail5):
+	mov	(%rsi), %ecx
+	mov	4(%rsi), %dx
+	mov	%ecx, (%rdi)
+	mov	%dx, 4(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail6):
+	mov	(%rsi), %ecx
+	mov	3(%rsi), %edx
+	mov	%ecx, (%rdi)
+	mov	%edx, 3(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail7):
+	mov	(%rsi), %rdx
+	mov	%rdx, (%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail8):
+	mov	(%rsi), %rcx
+	mov	8(%rsi), %dl
+	mov	%rcx, (%rdi)
+	mov	%dl, 8(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail9):
+	mov	(%rsi), %rcx
+	mov	8(%rsi), %dx
+	mov	%rcx, (%rdi)
+	mov	%dx, 8(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail10):
+	mov	(%rsi), %rcx
+	mov	7(%rsi), %edx
+	mov	%rcx, (%rdi)
+	mov	%edx, 7(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail11):
+	mov	(%rsi), %rcx
+	mov	8(%rsi), %edx
+	mov	%rcx, (%rdi)
+	mov	%edx, 8(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail12):
+	mov	(%rsi), %rcx
+	mov	5(%rsi), %rdx
+	mov	%rcx, (%rdi)
+	mov	%rdx, 5(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail13):
+	mov	(%rsi), %rcx
+	mov	6(%rsi), %rdx
+	mov	%rcx, (%rdi)
+	mov	%rdx, 6(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail14):
+	mov	(%rsi), %rcx
+	mov	7(%rsi), %rdx
+	mov	%rcx, (%rdi)
+	mov	%rdx, 7(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail15):
+	movdqu	(%rsi), %xmm0
+	movdqu	%xmm0, (%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail16):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %cl
+	movdqu	%xmm0, (%rdi)
+	mov	%cl, 16(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail17):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %cx
+	movdqu	%xmm0, (%rdi)
+	mov	%cx, 16(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail18):
+	movdqu	(%rsi), %xmm0
+	mov	15(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	mov	%ecx, 15(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail19):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	mov	%ecx, 16(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail20):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %ecx
+	mov	20(%rsi), %dl
+	movdqu	%xmm0, (%rdi)
+	mov	%ecx, 16(%rdi)
+	mov	%dl, 20(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail21):
+	movdqu	(%rsi), %xmm0
+	mov	14(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	mov	%rcx, 14(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail22):
+	movdqu	(%rsi), %xmm0
+	mov	15(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	mov	%rcx, 15(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail23):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %rcx
+	movdqu	%xmm0, (%rdi)
+	mov	%rcx, 16(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail24):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %rdx
+	mov	24(%rsi), %cl
+	movdqu	%xmm0, (%rdi)
+	mov	%rdx, 16(%rdi)
+	mov	%cl, 24(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail25):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %rdx
+	mov	24(%rsi), %cx
+	movdqu	%xmm0, (%rdi)
+	mov	%rdx, 16(%rdi)
+	mov	%cx, 24(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail26):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %rdx
+	mov	23(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	mov	%rdx, 16(%rdi)
+	mov	%ecx, 23(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail27):
+	movdqu	(%rsi), %xmm0
+	mov	16(%rsi), %rdx
+	mov	24(%rsi), %ecx
+	movdqu	%xmm0, (%rdi)
+	mov	%rdx, 16(%rdi)
+	mov	%ecx, 24(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail28):
+	movdqu	(%rsi), %xmm0
+	movdqu	13(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 13(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail29):
+	movdqu	(%rsi), %xmm0
+	movdqu	14(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 14(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail30):
+	movdqu	(%rsi), %xmm0
+	movdqu	15(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 15(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail31):
+	movdqu	(%rsi), %xmm0
+	movdqu	16(%rsi), %xmm2
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 16(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail32):
+	movdqu	(%rsi), %xmm0
+	movdqu	16(%rsi), %xmm2
+	mov	32(%rsi), %cl
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 16(%rdi)
+	mov	%cl, 32(%rdi)
+	RETURN
+
+	.p2align 4
+L(StringTail33):
+	movdqu	(%rsi), %xmm0
+	movdqu	16(%rsi), %xmm2
+	mov	32(%rsi), %cl
+	movdqu	%xmm0, (%rdi)
+	movdqu	%xmm2, 16(%rdi)
+	mov	%cl, 32(%rdi)
+	RETURN
+
+	.p2align 4
+L(CalculateSrcLenCase1):
+	xor	%r8, %r8
+	xor	%rax, %rax
+L(CalculateSrcLen):
+	pxor	%xmm0, %xmm0
+	xor	%rcx, %rcx
+	add	%r8, %rsi
+	movdqu	(%rsi), %xmm1
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+	test	%rdx, %rdx
+	jnz	L(SrcLenLoopEnd)
+
+	add	%rax, %r9
+	mov	$16, %rax
+	mov	%rsi, %rcx
+	and	$15, %rcx
+	and	$-16, %rsi
+L(SrcLenLoop):
+	movdqa	(%rsi, %rax), %xmm1
+	pcmpeqb	%xmm1, %xmm0
+	pmovmskb %xmm0, %rdx
+	test	%rdx, %rdx
+	jnz	L(SrcLenLoopEnd)
+	add	$16, %rax
+	jmp	L(SrcLenLoop)
+
+	.p2align 4
+L(SrcLenLoopEnd):
+	bsf	%rdx, %rdx
+	add	%rdx, %rax
+	sub	%rcx, %rax
+	RETURN
+
+END (STRLCPY)
+
+	.p2align 4
+	.section .rodata
+L(ExitTable):
+	.int	JMPTBL(L(Exit0), L(ExitTable))
+	.int	JMPTBL(L(Exit1), L(ExitTable))
+	.int	JMPTBL(L(Exit2), L(ExitTable))
+	.int	JMPTBL(L(Exit3), L(ExitTable))
+	.int	JMPTBL(L(Exit4), L(ExitTable))
+	.int	JMPTBL(L(Exit5), L(ExitTable))
+	.int	JMPTBL(L(Exit6), L(ExitTable))
+	.int	JMPTBL(L(Exit7), L(ExitTable))
+	.int	JMPTBL(L(Exit8), L(ExitTable))
+	.int	JMPTBL(L(Exit9), L(ExitTable))
+	.int	JMPTBL(L(Exit10), L(ExitTable))
+	.int	JMPTBL(L(Exit11), L(ExitTable))
+	.int	JMPTBL(L(Exit12), L(ExitTable))
+	.int	JMPTBL(L(Exit13), L(ExitTable))
+	.int	JMPTBL(L(Exit14), L(ExitTable))
+	.int	JMPTBL(L(Exit15), L(ExitTable))
+	.int	JMPTBL(L(Exit16), L(ExitTable))
+	.int	JMPTBL(L(Exit17), L(ExitTable))
+	.int	JMPTBL(L(Exit18), L(ExitTable))
+	.int	JMPTBL(L(Exit19), L(ExitTable))
+	.int	JMPTBL(L(Exit20), L(ExitTable))
+	.int	JMPTBL(L(Exit21), L(ExitTable))
+	.int	JMPTBL(L(Exit22), L(ExitTable))
+	.int	JMPTBL(L(Exit23), L(ExitTable))
+	.int	JMPTBL(L(Exit24), L(ExitTable))
+	.int	JMPTBL(L(Exit25), L(ExitTable))
+	.int	JMPTBL(L(Exit26), L(ExitTable))
+	.int	JMPTBL(L(Exit27), L(ExitTable))
+	.int	JMPTBL(L(Exit28), L(ExitTable))
+	.int	JMPTBL(L(Exit29), L(ExitTable))
+	.int	JMPTBL(L(Exit30), L(ExitTable))
+	.int	JMPTBL(L(Exit31), L(ExitTable))
+	.int	JMPTBL(L(Exit32), L(ExitTable))
+L(ExitStringTailTable):
+	.int	JMPTBL(L(StringTail0), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail1), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail2), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail3), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail4), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail5), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail6), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail7), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail8), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail9), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail10), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail11), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail12), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail13), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail14), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail15), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail16), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail17), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail18), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail19), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail20), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail21), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail22), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail23), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail24), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail25), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail26), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail27), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail28), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail29), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail30), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail31), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail32), L(ExitStringTailTable))
+	.int	JMPTBL(L(StringTail33), L(ExitStringTailTable))